A useful patch?

[Mike Stump]

> Date: 10 Apr 2000 14:53:22 -0000
> From: Andrew Haley <aph@pasanda.cygnus.co.uk>
> To: gcc@gcc.gnu.org

> I wrote tons of tracing code and eventually found the problem by
> adding a trap that triggers if the exceptioion handler chain is popped
> by the wrong function (any function other than the one that did the
> corresponding DHC push).

> This patch is potentially useful to anyone with problems in this area.
> But what should I do with it?

Submit it to gcc-patches.

> If I wrap the code in #ifdef SJLJ_EXECPTION_DEBUG or somesuch is it
> likely to be acceptable?
  
Yes, please wrap it.

> I've appended a rough (and untested) version of what I have in mind.
> If this sort of thing is acceptable I'll finish it up and submit it.

As the author of the code you had to debug, I am in favor of it
(wrapped and default off).

I had a function to trace out the data structures and print them, was
useful in debugging.